<p>One of the significant applications of AI tools in concrete technology is the prediction of the concrete’s compressive strength where it is tedious to follow the standard mix-design procedure. It is necessary for the researchers to identify the root cause for the lack of accuracy during and practical application. As per the scrupulous literature survey, two significant issues need to be addressed in predicting the properties of fibre reinforced concrete i.e., the data availability of those special concrete is less and disregard of special properties (fibre dosage, tensile strength of fibre) as inputs. To address these issues, three AI models were considered viz., artificial neural networks (deep learning model), support vector machine (data driven model) and random forest regression (regression model) to anticipate the compressive strength of concrete with carbon and glass fibers. The rationale for the selection of AI models depends on the selection of 3 different algorithms which offer adaptability, flexibility and ability to handle complex problems. It is expected that the AI models should have the ability to assess the sensitive input parameter (fibre dosage) and predict the strength with small database. To ensure this, a detailed comparative analysis was made. As per R² values, neural network, support vector and random forest models showed 90%, 83% and 84% respectively. Among the three models, neural network has the better capability to recognize the sensitive parameter as the prediction curve is similar to experimental curve with small margin of residue. The other two models can predict the accurate strength upto 50% of the data sets however accompanied with poor recognition of sensitive parameters and anomalous results.</p>